Abstract
Our primary objective is to verify or refute a 2013 study by Connolly et al. which showed that in early pregnancy, a gestational sac was visualized 99% of the time on transvaginal ultrasound when the HCG level reached 3510 mIU/mL. Our secondary objective was to make clinical correlations by assessing the relationship between human chorionic gonadotropin (HCG) level in early pregnancy when a gestational sac is not seen and pregnancy outcomes of live birth, spontaneous abortion, and ectopic pregnancy. This retrospective study includes 144 pregnancies with an outcome of live birth, 87 pregnancies with an outcome of spontaneous abortion, and 59 ectopic pregnancies. Logistic regression is used to determine the probability of visualizing a gestational sac and/or yolk sac based on the HCG level. A gestational sac is predicted to be visualized 50% of the time at an HCG level of 979 mIU/mL, 90% at 2421 mIU/mL, and 99% of the time at 3994 mIU/mL. A yolk sac was predicted to be visualized 50% of the time at an HCG level of 4626 mIU/mL, 90% at 12,892 mIU/mL, and 99% at 39,454 mIU/mL. A total of 90% of ectopic pregnancies presented with an HCG level below 3994 mIU/mL. These results are in agreement with the study by Connolly et al. Since most early ectopic pregnancies had an HCG value below the discriminatory level for gestational sac visualization, other methods for the evaluation of pregnancy of unknown location such as repeat HCG values are clinically important.